当前位置:网站首页>Alibaba cloud microservices (IV) service mesh overview and instance istio
Alibaba cloud microservices (IV) service mesh overview and instance istio
2022-07-06 13:09:00 【Geometer】
To mention Service Mesh I have to mention microservices , According to Wikipedia
Microservices (Microservices) Is a software architecture style , It is a small functional block focusing on a single responsibility and function (Small Building
Blocks) Based on , Using modularity to compose complex large applications , The use of each functional block has nothing to do with language
(Language-Independent/Language agnostic) Of API Sets communicate with each other
With Google three carriages BIGTABLE,Mapreduce,GFS Appearance , Knock on the door of distributed , Fusing strategy 、 Load balancing 、 The emergence of service discovery , Services have some communication semantics according to business needs , In order to avoid each service developing its own set of communication semantics , There is a micro service framework , for instance Spring Cloud Other framework , They realize the semantic functions required by the distributed system , Like service discovery , Load balancing, etc. , To some extent, the communication details are shielded , So that developers can realize the whole function with less code without considering some problems of communication .
But later , People also found that the microservice framework is not omnipotent , It mainly has the following problems :
1. Developers need to spend a lot of time learning some details of the microservice framework , Because if something goes wrong , It is difficult to troubleshoot and solve problems without understanding the framework
2. Frameworks that use microservices need to use specific languages , This is different from the original features of microservices ---- It has nothing to do with language
To solve this problem , first generation Service Mesh emerge as the times require
It abstracts the communication of distributed system into a separate layer , It is up to him to realize load balancing 、 Service discovery 、 Certificate authority 、 Surveillance tracking 、 Flow control and other functions required by distributed systems , As a proxy service equivalent to a service , Deployed with services , It completes the request between services through the communication between agents .
Pictured , The blue part is the service grid , If you only look at the blue part , It's easy to understand why it's called “Mesh” Is that right ~
later , For unified management , People have derived the second generation Service Mesh, Developed a centralized control panel , All stand-alone components update the network topology strategy and report the data through the control panel , A very classic example is Istio
Istio It consists of two parts : Control plane and data plane .
The data plane is the communication plane between businesses . If there is no service grid , The network cannot understand the traffic being sent , Nor can it be based on what type of traffic it is , Or who it comes from , To whom to make any decision .
The service grid uses agents to intercept all network traffic , Allows you to provide a wide range of application aware functions according to the configuration you set .
The agent is deployed with each service you start in the cluster , Or run with a service running on a virtual machine .
The control plane gets the configuration and service views you need , And dynamically program the proxy server , Update them as rules or environments change .
Use Istio Before
Use Istio after :
Service Mesh Advantages and disadvantages :
advantage :
1. Shielding distributed communication complexity
2. Shielding language
3. Transparent to application layer , Don't need to care about
shortcoming :
1. The communication performance is reduced to a certain extent
2. The overall stability of the service depends on Service Mesh
边栏推荐
- 初识C语言(上)
- 如何保障 MySQL 和 Redis 的数据一致性?
- 【rtklib】在rtk下使用抗差自适应卡尔曼滤波初步实践
- MySQL 30000 word essence summary + 100 interview questions, hanging the interviewer is more than enough (Collection Series
- [algorithm] sword finger offer2 golang interview question 6: sum of two numbers in the sorting array
- [algorithm] sword finger offer2 golang interview question 7: 3 numbers with 0 in the array
- Ten minutes to thoroughly master cache breakdown, cache penetration, cache avalanche
- Novatel board oem617d configuration step record
- [algorithm] sword finger offer2 golang interview question 4: numbers that appear only once
- 记录:动态Web项目servlet访问数据库404错误之解决
猜你喜欢
[algorithme] swordfinger offer2 golang question d'entrevue 2: addition binaire
[算法] 剑指offer2 golang 面试题5:单词长度的最大乘积
[算法] 剑指offer2 golang 面试题6:排序数组中的两个数字之和
[算法] 剑指offer2 golang 面试题4:只出现一次的数字
阿里云微服务(三)Sentinel开源流控熔断降级组件
十分钟彻底掌握缓存击穿、缓存穿透、缓存雪崩
面试必备:聊聊分布式锁的多种实现!
[算法] 剑指offer2 golang 面试题3:前n个数字二进制形式中1的个数
几道高频的JVM面试题
How to ensure data consistency between MySQL and redis?
随机推荐
GNSS positioning accuracy index calculation
[algorithm] sword finger offer2 golang interview question 12: the sum of the left and right sub arrays is equal
[rtklib] preliminary practice of using robust adaptive Kalman filter under RTK
TYUT太原理工大学2022数据库大题之概念模型设计
Role movement in the first person perspective
错误:排序与角标越界
[algorithme] swordfinger offer2 golang question d'entrevue 2: addition binaire
On March 15, the official version of go 1.18 was released to learn about the latest features and usage
抽象类和接口
Fundamentals of UD decomposition of KF UD decomposition [1]
Answer to "software testing" exercise: Chapter 1
Rt-ppp test using rtknavi
TYUT太原理工大学2022数据库大题之E-R图转关系模式
[algorithm] sword finger offer2 golang interview question 4: numbers that appear only once
Problems and solutions of robust estimation in rtklib single point location spp
KF UD decomposition pseudo code implementation advanced [2]
What are the advantages of using SQL in Excel VBA
一文搞定 UDP 和 TCP 高频面试题!
Branch and loop statements
Excel导入,导出功能实现